Sector guides
The EU AI Act, by sector
The Act does not name industries — it names uses. Annex III lists eight areas, Article 5 lists prohibited practices, Article 50 lists disclosure duties, and whether any of it reaches you depends on which of those uses you happen to run. These guides work backwards from the sector to the provisions, and quote them.
Each one covers the same ground: what the Act names in that sector, how its common use cases classify, and what gets misread — in both directions, since treating something as high-risk when the Act carves it out is the more expensive mistake.
HR and recruitment
Recruitment is the most explicitly covered commercial use in the Act. Annex III point 4 names it directly, which means most hiring tools are high-risk by description rather than by argument — and one common HR use is prohibited outright.
Financial services and insurance
Two points of Annex III do most of the work here, and both contain carve-outs that decide whether a system is in scope at all. Reading them loosely is how banks end up treating fraud models as high-risk and insurers treating motor pricing as caught.
Healthcare and life sciences
Most clinical AI is high-risk through the product-safety route in Article 6(1), not through Annex III — which means the classification turns on device law rather than on how clinical the use feels. Annex III adds two further hooks for public services and emergency response.
Education and training
Annex III point 3 names four distinct education uses, covering the whole path from admission to assessment to invigilation. As in the workplace, inferring emotion is prohibited rather than regulated.
Public sector and law enforcement
The densest part of the Act. Four Annex III areas and most of the Article 5 prohibitions are aimed at public authorities, and several of the bans apply to law enforcement specifically rather than to everyone.
Retail, marketing and media
Little in retail is high-risk — recommendation and personalisation are not named in Annex III at all. What does bite here are the manipulation prohibitions and, from the 2026 amendment onward, the duty to mark synthetic content.
Manufacturing and infrastructure
Industrial AI is reached through the product-safety route rather than Annex III — and this is the sector the 2026 amendment changed most, by narrowing what counts as a safety component and moving machinery to a sectoral approach.
Technology and software vendors
If you build the model or ship the system, obligations attach to you as provider — and general-purpose models carry their own regime in Articles 53 and 55, separate from the high-risk tiers entirely.